fix: dev-orchestrator Copilot CLI compatibility (#3010, #3011)

.claude/skills/dev-orchestrator/SKILL.md

@@ -95,12 +95,15 @@ that instructs Claude to invoke `Skill(skill="dev-orchestrator")` for dev/invest
 
 ## Execution Instructions
 
-**YOUR NEXT ACTION after reading this skill MUST be a Bash tool call that
-executes `run_recipe_by_name("smart-orchestrator")`.**
+**YOUR NEXT ACTION after reading this skill MUST include a Bash tool call that
+executes `run_recipe_by_name("smart-orchestrator")`.** If your runtime requires
+a `report_intent` call (e.g. Copilot CLI), emit it **in parallel** with the
+Bash launch — both tool calls in the same response. The constraint is that no
+*other* substantive action (reading files, calling TodoWrite, invoking Agent,
+or typing a response) may precede or replace the Bash launch.
 
-If you find yourself doing anything else — reading more files, calling
-TodoWrite, invoking Agent, or typing a response — you are bypassing the
-workflow. Stop. Go back. Run the recipe runner.
+If you find yourself doing anything else instead of launching the recipe
+runner, you are bypassing the workflow. Stop. Go back. Run the recipe runner.
 
 Development and Investigation tasks have exactly ONE execution path: the
 recipe runner. There is no "Option B." There is no "I'll follow the steps
@@ -116,7 +119,8 @@ When this skill is activated:
 
 ### REQUIRED: Execute via Recipe Runner — IMMEDIATELY
 
-Your very next tool call must launch the recipe runner. **CRITICAL**: Claude
+Your next tool call(s) must include the recipe runner launch (alongside
+`report_intent` if your runtime requires it). **CRITICAL**: Claude
 Code's `run_in_background` kills processes after ~10 minutes. Recipe
 workstreams can take hours. You MUST use a tmux session for execution:
 
@@ -145,6 +149,24 @@ print(f'Recipe result: {result}')
 - `tmux new-session -d` — detached session, no timeout, survives disconnects
 - Monitor with: `tail -f /tmp/recipe-runner-output.log` or `tmux attach -t recipe-runner`
 
+**Restarting a stale tmux session**: Some runtimes (e.g. Copilot CLI) block
+`tmux kill-session` because it does not target a numeric PID. Use one of these
+shell-policy-safe alternatives instead:
+
+```bash
+# Option A (preferred): use a unique session name per run to avoid collisions
+tmux new-session -d -s "recipe-$(date +%s)" "..."
+
+# Option B: locate the tmux server PID and terminate with numeric kill
+tmux list-sessions -F '#{pid}' 2>/dev/null | xargs -I{} kill {}
+
+# Option C: let tmux itself handle it — send exit to all panes
+tmux send-keys -t recipe-runner "exit" Enter 2>/dev/null; sleep 1
+```
+
+If using Option A, update the `tail -f` / `tmux attach` commands to use the
+same session name.
+
 **DO NOT use `run_in_background`** for recipe execution — it will be killed
 after ~10 minutes (Issue #2909).
 

pyproject.toml

@@ -5,7 +5,7 @@ backend-path = ["."]
 
 [project]
 name = "amplihack"
-version = "0.6.16"
+version = "0.6.17"
 description = "Amplifier bundle for agentic coding with comprehensive skills, recipes, and workflows"
 requires-python = ">=3.11"
 dependencies = [

tests/skills/test_dev_orchestrator_issue_3010_3011.py

@@ -0,0 +1,63 @@
+"""Regression tests for issues #3010 and #3011.
+
+Issue #3010: dev-orchestrator must allow report_intent in parallel with the
+Bash launch call, not hard-require Bash as the *only* next tool call.
+
+Issue #3011: dev-orchestrator must provide shell-policy-safe tmux restart
+guidance that does not rely on ``tmux kill-session``.
+"""
+
+from pathlib import Path
+
+REPO_ROOT = Path(__file__).resolve().parents[2]
+DEV_ORCHESTRATOR_SKILL = REPO_ROOT / ".claude" / "skills" / "dev-orchestrator" / "SKILL.md"
+
+
+class TestIssue3010ReportIntentCompatibility:
+    """The skill must not hard-require Bash as the sole next tool call."""
+
+    def setup_method(self):
+        self.content = DEV_ORCHESTRATOR_SKILL.read_text()
+
+    def test_no_hard_must_be_a_bash_tool_call(self):
+        """The old phrasing 'MUST be a Bash tool call' is replaced."""
+        assert "MUST be a Bash tool call that" not in self.content
+
+    def test_allows_report_intent_in_parallel(self):
+        """The skill explicitly mentions report_intent as acceptable."""
+        assert "report_intent" in self.content
+
+    def test_must_include_bash_launch(self):
+        """The skill still requires a Bash launch — just not as the *sole* call."""
+        assert "MUST include a Bash tool call" in self.content
+
+    def test_next_tool_calls_plural_form(self):
+        """The follow-up instruction uses the plural 'tool call(s)'."""
+        assert "next tool call(s)" in self.content
+
+
+class TestIssue3011TmuxRestartGuidance:
+    """The skill must provide shell-policy-safe tmux restart alternatives."""
+
+    def setup_method(self):
+        self.content = DEV_ORCHESTRATOR_SKILL.read_text()
+
+    def test_no_kill_session_as_primary_instruction(self):
+        """The skill must not recommend ``tmux kill-session`` as the restart path."""
+        assert "tmux kill-session -t recipe-runner" not in self.content
+
+    def test_unique_session_name_option(self):
+        """Option A: unique session name per run."""
+        assert "recipe-$(date +%s)" in self.content
+
+    def test_numeric_pid_option(self):
+        """Option B: numeric PID termination."""
+        assert "#{pid}" in self.content
+
+    def test_send_exit_option(self):
+        """Option C: send exit to panes."""
+        assert 'send-keys -t recipe-runner "exit"' in self.content
+
+    def test_shell_policy_safe_section_exists(self):
+        """A dedicated section on restarting stale tmux sessions exists."""
+        assert "Restarting a stale tmux session" in self.content